How Gains in Learning Disability Knowledge Enhance Pre-Service Teachers’ Self-Efficacy Through Attitudinal Shifts

Abstract

Recent studies indicate that Chinese K-12 teachers possess insufficient knowledge regarding learning disability (LD), hindering their ability to provide effective instruction. Given the foundational role of the pre-service phase in cultivating a scientific approach to teaching, this study aimed to boost pre-service teachers’ LD knowledge and explore its subsequent impact on teaching efficacy and attitudes. Fifty-one pre-service teachers with low levels of baseline LD knowledge were randomly assigned to either a training group or a control group. Utilizing a pretest–intervention–posttest design, the study measured changes in LD knowledge, teaching efficacy, and attitudes toward students with LD. Crucially, attitudes were assessed via a vignette paradigm that differentiated between two components of cognitive evaluations (expectations of future student failure) and emotional experiences (anger arousal towards academic failure). The results showed that pre-service teachers in the training group exhibited substantial gains in LD knowledge. These knowledge gains significantly predicted enhanced teaching efficacy, but this relationship was indirect. Mediation analysis revealed that improved knowledge reduced anger arousal, which in turn boosted efficacy. These findings suggest that fostering teaching confidence requires more than mere knowledge accumulation; it also entails using LD-related knowledge to mitigate negative emotions toward struggling learners. This underscores that teacher education programs must incorporate explicit cultivation of emotional and attitudinal competencies alongside conventional cognitive training.

1. Introduction

Learning disabilities (LDs) are neurodevelopmental conditions characterized by persistent difficulties in acquiring academic skills, despite adequate intelligence, conventional instruction, and socio-cultural opportunities (World Health Organization 2019). Globally, the prevalence of LDs remains relatively stable, estimated at approximately 8–9% across countries (Y. Li et al. 2023; Y. Q. Li et al. 2025; Scaria et al. 2023). For instance, data obtained from the 1997 to 2021 National Health Interview Survey (i.e., a nationally representative household survey using multistage stratified sampling) reported an 8.83% prevalence rate of LDs among U.S. children and adolescents aged 6–17 years (Y. Li et al. 2023). A meta-analysis reported a prevalence rate of about 8% among children up to 19 years in India (Scaria et al. 2023). In mainland China, a prevalence rate of 8.9% has been reported among children aged 7–13 years (Y. Q. Li et al. 2025). Typically identified in early developmental stages, learning disabilities can endure throughout an individual’s lifespan. These disabilities mainly encompass dyslexia, characterized by significant and persistent difficulties in accurate word recognition, decoding, and spelling (World Health Organization 2019); dysgraphia, which manifests as impairments in letter formation/legibility, letter spacing, spelling, fine motor coordination, writing fluency, grammar, and composition (Chung et al. 2020); and dyscalculia, a specific learning disability in symbolic number processing and calculation (Decarli et al. 2020). A frequent comorbidity is attention-deficit/hyperactivity disorder (ADHD), which is defined by impairing levels of inattention, disorganization, and/or hyperactivity-impulsivity (American Psychiatric Association 2022). The co-occurrence of ADHD exacerbates learning challenges by further impairing attention, impulse control, and executive functioning (Liu et al. 2024; Z. Zhang et al. 2025).

Students with learning disabilities often exhibit lower academic attainment compared to their typically developing peers, yet the majority are educated in inclusive mainstream settings (Bruefach and Reynolds 2022; Daniel 2025; Sáenz and Fuchs 2002). A critical barrier to effective inclusion is the inaccurate identification of LDs by educators, which delays necessary interventions (Stone et al. 2023). This challenge is compounded by a dearth of appropriate pedagogical strategies, leading to a lack of confidence among educators regarding the delivery of effective instructional support (Westphal et al. 2024). A pivotal factor contributing to this issue is the inadequacy of scientific knowledge concerning LDs. Drawing upon established literature (Soriano-Ferrer et al. 2016; Yin et al. 2020), educators’ scientific knowledge of LDs is conceptualized as a multidimensional construct. It encompasses a theoretical understanding of the etiology, nature, and outcomes of LDs, alongside a grasp of their cognitive, behavioral, and biological underpinnings. Furthermore, this knowledge base extends to practical application, including the identification of at-risk students to facilitate timely intervention, as well as the implementation of appropriate diagnostic and instructional strategies. Given that current educators often lack sufficient scientific knowledge of LDs and possess a limited understanding of the relationship between brain development and educational practices (Ching et al. 2020), and considering that the construction of professional knowledge primarily occurs during teacher education programs (Y. Zhang et al. 2024), the present study targeted pre-service teachers. An online course was developed to enhance pre-service teachers’ understanding of LDs, covering definitions, possible causes, identification, psychological characteristics, and evidence-based educational strategies for students with LDs. This study aimed to examine how the enhancement of such targeted knowledge dynamically influences their teaching performance.

1.1. Teachers’ Limited Knowledge of Learning Disabilities

A disconnect has been observed between educational practice and scientific evidence regarding LDs. Early research by Allington (1982) identified a persistent and specific misconception among educators, that is, the adherence to the visual perceptual deficit hypothesis. Their study found that 40% to 70% of surveyed teachers erroneously believed that reading difficulties were primarily caused by visual-perceptual issues such as letter reversals rather than deficits in phonological processing. This misattribution led to the endorsement of ineffective remediation strategies, such as visual training (e.g., eye exercises or block designs), despite empirical evidence demonstrating their inefficacy. Although research into the mechanisms of LDs has progressed, misconceptions regarding their etiology and intervention still persist globally among teachers (Gyasi et al. 2020; Knight 2018; Poznanski et al. 2018; Sousa et al. 2017; Washburn et al. 2011, 2014; White et al. 2020; Williams 2013; Yin et al. 2020; Yüksel and Özcan 2025). For instance, many educators fail to accurately distinguish between LDs, intellectual disability, and general academic underachievement. They often attribute LDs to intellectual disability or inadequate educational support which are explicitly excluded by the defining criteria for LDs (Vaughn and Fuchs 2003; Soriano-Ferrer et al. 2016). Additionally, educators frequently overlook the neurophysiological basis of LDs (Knight 2018; Washburn et al. 2014), erroneously believing that children can outgrow LDs, rather than recognizing them as lifelong neurodevelopmental disorders (Washburn et al. 2014).

Notably, regional disparities exist in teachers’ knowledge of LDs. In mainland China, relevant studies indicate that primary and secondary school teachers demonstrate a particularly limited scientific understanding of these conditions (Ching et al. 2020; Hu 2005; Yin et al. 2020). The study of Hu (2005) found that less than 60% of teachers in Beijing could identify appropriate interventions for these students. Comparative data further highlights this gap. Using similar questionnaires on LD knowledge, researchers reported that pre-service teachers in the United States and the United Kingdom, as well as LD experts in Australia demonstrated accuracy ranging from approximately 71% to 88% (Serry and Hammond 2015; Washburn et al. 2014). In contrast, in-service teachers in developed regions of mainland China scored merely 54% on similar assessments (Yin et al. 2020). While the overall knowledge deficit is evident, a comparative analysis of specific misconceptions reveals both cross-cultural consistencies and distinct cultural differences. A universal misconception is the belief that dyslexia (i.e., one of the main types of LDs) stems from visual deficits, leading to the erroneous endorsement of visual interventions like colored overlays (Washburn et al. 2014; Yin et al. 2020). However, specific knowledge gaps differ significantly. While Western teachers correctly reject environmental causes, a substantial proportion of Chinese teachers (approximately 60%) erroneously attribute dyslexia to a literacy-poor home environment (Yin et al. 2020). Furthermore, unlike their Western counterparts who often mistakenly view LD as a temporary developmental stage, Chinese teachers predominantly lack an understanding of the neurobiological etiology of LDs. This indicates that while misconceptions about the behavioral manifestations of dyslexia are universal, the specific deficits in understanding the biological and cognitive foundations are more severe among Chinese educators (Yin et al. 2020). This knowledge gap is often accompanied by the endorsement of neuromyth, which refers to a misconception generated by misunderstanding, misreading, or misquoting scientific facts established by brain research to judge the use of such research in educational and other contexts (Organization for Economic Co-Operation and Development 2002). Further analysis by Yin et al. (2020) highlighted significant misconceptions among Chinese teachers regarding the neurophysiological mechanisms of LDs. For instance, many hold the erroneous belief that learning problems associated with developmental differences in brain function cannot be remediated through education. Similarly, studies on pre-service teachers and in-service inclusive education teachers in Hong Kong revealed a limited understanding of the relationship between brain development and educational practices, manifesting as misconceptions akin to neuromyths (Ching et al. 2020; Tsang et al. 2024). This issue extends beyond China; a global trend suggests that insufficient neuroscience knowledge can lead teachers to embrace pseudoscientific practices. In a review of 24 studies, Torrijos-Muelas et al. (2021) found that teachers from at least ten countries (mainly in Europe and America) generally agreed with around 40–80% of the neuromyth statements. This suggests that while the specific myths may vary, the underlying vulnerability to pseudoscientific beliefs due to insufficient training remains a critical challenge for educator preparation both locally and internationally.

1.2. Teacher Training on Knowledge About Learning Disabilities

Insufficient knowledge could result in the neglect of learners with special needs, thereby impeding their global development (Buli-Holmberg and Jeyaprathaban 2016). To overcome this bottleneck, teacher education has proven effective in enhancing educators’ understanding of LDs. Such interventions are critical for equipping teachers with the skills to establish supportive classroom environments and devise instructional strategies to manage challenging behaviors (Beck et al. 2026; Berryhill et al. 2025; Ward et al. 2022).

Previous research indicates that training programs can effectively enhance teachers’ theoretical and practical knowledge regarding LD (e.g., Aguiar et al. 2014; Alabd et al. 2018; Barnett et al. 2012). Early studies, such as Barnett et al. (2012) implemented a 7-week web-based intervention among primary school teachers, incorporating web links, emails, and discussion forums focused on LDs (mainly ADHD). This study found that teachers reported significantly improved perceived mastery of ADHD knowledge following the intervention as compared to baseline.

Subsequent researchers employed standardized LD knowledge tests to assess theoretical mastery across three domains including etiology and impact, symptoms and diagnosis, and intervention strategies (Aguiar et al. 2014; Alabd et al. 2018). Aguiar et al. (2014) implemented a one-day training totaling 320 min for primary school teachers. This training program incorporated clinical vignettes addressing frequent misconceptions about LDs, lectures on LD symptoms and etiology, as well as strategies for managing LDs in the school setting. Results indicated that the training effectively reduced misconceptions regarding etiology and teaching strategies, and increased the acceptability of classroom behavior management techniques. Similarly, Alabd et al. (2018) administered a brief one-day training consisting of four 30–35 min sessions (totaling approximately 2 h), covering definitions, causes, diagnosis, and management techniques of LDs. This study reported that the brief intervention significantly improved teachers’ knowledge levels across all three domains of LD knowledge.

Furthermore, the benefits of such training demonstrate a degree of retention. Syed and Hussein (2010) conducted a 5-day training program (totaling 10 h) for primary and secondary teachers. The training covered psychoeducational development, DSM-based diagnostic criteria, comorbidity, and practical behavior techniques through case scenarios. Relevant results showed a significant increase in LD knowledge at post-test compared to pre-test, with training gains remaining significant at a 6-month follow-up.

However, it should be noted that most of the aforementioned studies relied solely on pre-post comparisons within a single training group, lacking a control group; consequently, relevant results may be confounded by placebo effects (Hafliðadóttir et al. 2021). Later studies employing more rigorous designs have corroborated these findings. Jarque Fernández et al. (2021) employed a randomized control trial design where the intervention group underwent extensive training comprising 17 two-h sessions over nearly a year. This program included general knowledge, intervention techniques, behavioral strategies, classroom management, stress management, and communication skills regarding LDs. Results showed significant improvements across all knowledge domains for the training group, whereas the control group showed no significant changes. Similarly, Latouche and Gascoigne (2019) utilized a training-control group design to deliver a brief 2 h workshop to nearly 300 primary school teachers. The training covered etiology, neuropsychological impairments, assessment, cultural conceptualizations, treatment, and classroom management strategies. Results indicated that while the control group showed no pre-post differences, the training group exhibited significant gains in LD knowledge, which were retained at a 1-month follow-up.

Collectively, existing research demonstrates that short-term courses (ranging from 2 to 7 h) can remarkably improve primary school teachers’ knowledge levels and instructional management of LDs compared to baseline (Aguiar et al. 2014; Alabd et al. 2018). This confirms the direct intervention effects on the specific domains covered by the training and supports the feasibility of implementing such interventions in school settings. However, a significant challenge remains regarding the scope of current training, as most existing studies have concentrated on ADHD-related teacher training, leaving a notable paucity of research focusing on more prototypical learning difficulties, such as dyslexia and dyscalculia. Addressing this gap is a primary consideration in the design of the teacher training curriculum for the present study.

1.3. Impact of LD Knowledge Training on Teaching Efficacy and Teachers’ Attitudes Towards Students with Learning Disabilities

Previous research on LD knowledge training has primarily investigated its transfer effects on teaching efficacy (Alabd et al. 2018; Jarque Fernández et al. 2021; Latouche and Gascoigne 2019). Empirical studies have consistently demonstrated that both long-term and short-term teacher education programs of LD knowledge significantly enhance primary school teachers’ self-efficacy (Alabd et al. 2018; Jarque Fernández et al. 2021; Latouche and Gascoigne 2019), with positive effects remaining significant at a 1-month follow-up (Latouche and Gascoigne 2019). These training programs typically encompassed comprehensive content, including theoretical foundations (e.g., etiology, diagnosis, and neuropsychological impairments), practical classroom management strategies (e.g., behavioral interventions and instructional adaptations), and general professional skills (e.g., stress management and communication). Such contents constitute teachers’ general pedagogical and psychological knowledge regarding individual student characteristics. Given that this type of professional knowledge has been found to have a particular impact on classroom management and instructional quality (Voss and Kunter 2013), it provides a plausible explanation for the observed causal relationship between the acquisition of LD knowledge and the enhancement of teaching efficacy. Additionally, grounded in Bandura’s self-efficacy theory (1977), enactive mastery experience and vicarious experience are identified as primary sources that obviously influence efficacy. Enactive mastery experiences, considered the most influential source, refer to authentic evidence regarding one’s capability derived from personal successes and failures (Lippke 2017). Vicarious experience refers to the process of observing others (particularly peers or models perceived as similar) successfully performing (Rowland et al. 2020). It should be noted that within LD training, both knowledge instruction and practical operations serve to enhance enactive mastery experience. Meanwhile, case studies and group discussions contribute to building vicarious experience. This offers an alternative explanation for how increased LD knowledge boosts efficacy, that is, possibly by simultaneously fostering both mastery and vicarious experiences, the training ultimately promotes greater teaching efficacy.

Furthermore, some research has examined the possible impact of LD knowledge on teachers’ attitudes towards students with LDs. Relevant findings indicated that web-based interventions (i.e., 7-week programs incorporating discussion forums, email communication, and web link resources) inclined teachers to adopt more scientific attitudes toward LDs (Barnett et al. 2012). However, as this study relied on self-report attitude scales for measurement, there is a need for more objective assessment tools to clarify the effects of LD knowledge enhancement on teacher attitudes in future research. Moreover, it is important to recognize that attitude is a multidimensional construct comprising distinct components (Rohmer et al. 2022), such as the cognitive component (rational beliefs and evaluations) and the emotional component (emotional experiences and feelings). Previous studies utilizing subjective scales often measured these dimensions concurrently, capturing both teachers’ rational evaluations of the difficulty in educating students with LDs and their subjective emotional experiences during the process. Consequently, a critical question remains: Does the enhancement of LD knowledge promote improvements in both of these attitudinal components simultaneously, or does it elicit a more specific change in one particular component? This distinction warrants further investigation to be clearly established.

Relevant studies have typically examined the relationship between increased LD knowledge and teacher attitudes, as well as its link to teaching efficacy, in isolation. However, the multidimensional model of professional teaching competencies proposed by Baumert and Kunter (2013) underscores the necessity of comprehensively examining the interplay among LD knowledge, teacher attitudes, and teaching efficacy. This model delineates professional teaching competencies into two primary domains including cognitive and affective/motivational characteristics. Specifically, cognitive teacher characteristics refer to the teacher’s knowledge base, mental representations, and diagnostic reasoning regarding subject matter. Conversely, affective/motivational teacher characteristics pertain to teachers’ beliefs in their capabilities and their willingness to expend effort when facing challenges (Westphal et al. 2024). Within this framework, LD knowledge represents a core cognitive characteristic, while teaching attitude (especially its emotional component) serves as a key affective/motivational characteristic. Furthermore, teaching efficacy has been seen as an important aspect of professional competence that is closely associated with instructional behavior (Klassen et al. 2011; Li and Guo 2024; Tschannen-Moran and Hoy 2001). According to the cognitive reappraisal theory within the emotion regulation model (Gross 1999, 2002), modifying one’s cognitive interpretation of a situation can lead to more positive appraisals and effectively regulate emotional responses. Moreover, the emotional dimension of attitude corresponds to the efficacy source of physiological/affective states. Teachers who experience less anger when facing a student’s failure are in a more resourceful affective state, which may in turn support a higher level of self-efficacy. Therefore, a mediation model can be hypothesized, that is, LD knowledge may influence teaching efficacy indirectly by shaping teacher attitudes particularly its emotional dimension. Besides that, previous research indicated that increased LD knowledge contributed to improving the accuracy of teachers’ cognitive evaluations (Barnett et al. 2012). Crucially, the cognitive dimension of attitude acts as a perceptual lens through which enactive mastery experiences are appraised. Relevant research (Krischler and Pit-ten Cate 2019) indicated that teachers’ cognitive stereotypes, specifically perceptions of students’ competence, serve as a critical filter for processing student outcomes. When teachers hold preconceived cognitions that students with special needs possess low competence (Krischler and Pit-ten Cate 2019), these appraisals cast a shadow over the mastery experience. Consequently, even neutral or ambiguous student performance may be cognitively reconstructed as failure or a lack of progress. This cognitive reconstruction transforms potential mastery experiences into perceived threats, thereby undermining the foundation of teaching efficacy. Collectively, it is reasonable to infer that the enhancement of LD knowledge may not only positively influence the emotional component of teachers’ attitudes but also refine teachers’ cognitive evaluations of students with LD, ultimately fostering higher teaching efficacy. This theoretical linkage warrants further empirical clarification.

1.4. Aims of the Present Study

The present study utilized a pretest–intervention–posttest design to investigate the impact of LD knowledge training on the teaching efficacy of pre-service teachers in mainland China, while also examining the potential mediating role of teachers’ attitudes towards students with LDs. Previous research on teacher education (e.g., Alabd et al. 2018; Jarque Fernández et al. 2021; Latouche and Gascoigne 2019) has predominantly focused on in-service teachers. However, considering the substantial workload and time constraints faced by in-service educators, which often preclude opportunities for systematic learning, a shift in focus is warranted. Moreover, the pre-service period is a crucial stage for teachers to develop scientific teaching beliefs (Fessler 2006). These beliefs are intrinsically linked to teachers’ attitudes to students with special needs and their future teaching efficacy (Robertson et al. 2017). Consequently, targeting pre-service teachers is imperative for fostering optimal professional development. This study specifically focuses on mainland China, where K-12 teachers’ knowledge about LDs remains significantly lower compared to their counterparts in Europe and the United States (Yin et al. 2020). The training program comprised evidence-based content relevant to the future professional needs of teaching students with LDs. Furthermore, by measuring teachers’ attitudes towards students with LDs via a vignette paradigm that separately assessed cognitive and emotional dimensions, we aimed to explore the possible fine-grained mediating effect of these attitudes on the relationship between LD knowledge and teaching efficacy. Based on relevant literature (Lu et al. 2020; San Martin et al. 2021), we hypothesized that firstly, the training program would increase pre-service teachers’ LD knowledge, optimize their attitudes towards students with LDs, and boost their teaching efficacy. Secondly, and more importantly, drawing on the dynamic changes in LD knowledge, teacher attitudes, and teaching efficacy following LD knowledge training and the multidimensional model of professional teaching competencies proposed by Baumert and Kunter (2013), we predicted a relational model among the three variables. As shown in Figure 1, we hypothesized that training-induced gains in LD knowledge (i.e., cognitive teacher characteristic) would positively refine teachers’ attitudes (i.e., affective/motivational teacher characteristic). This refinement in attitudes would encompass both cognitive evaluations of students’ abilities and emotional arousal in response to LD-related situations, thereby fostering their teaching efficacy (i.e., a predictor associated with teaching behavior and competencies), which may be driven by changes in the information sources of mastery experience and physiological/affective states. Additionally, referring to relevant literature (Alabd et al. 2018; Jarque Fernández et al. 2021; Latouche and Gascoigne 2019), we also predicted a direct effect of increased LD knowledge on the enhancement of teaching efficacy, potentially involving changes in mastery and/or vicarious experiences.

2. Materials and Methods

2.1. Participants

The study protocol was approved by the Ethics Committee of [Corresponding Author’s Institution]. All procedures were conducted in accordance with the Declaration of Helsinki.

A total of 221 postgraduates were recruited from normal universities in Beijing (17 males, 204 females). All participants were specifically enrolled in Master of Education programs, which are professional degrees designed to prepare students for immediate employment as K-12 teachers upon graduation. To ensure a standardized level of professional competence, inclusion criteria required that all participants hold a valid teaching certification which students are eligible to apply for and awarded upon passing a national examination in mainland China, and have completed at least three months of educational practice. Based on the method of deviation analysis (Zhao et al. 2018), we screened 51 pre-service teachers from the initial pool whose scores in both LD knowledge and teaching efficacy were both lower than −1 standard deviation below the mean of the entire pool. These individuals were invited to participate in the teacher training project. Detailed information about the assessments of LD knowledge and teaching efficacy is introduced in Section 2.3. The 51 selected participants were then randomly assigned to either a training group (n = 26; 1 male, 25 females) or a control group (n = 25; 2 males, 23 females).

2.2. Procedures

There are three sessions in this intervention study: pre-test, training, and post-test. During the pre-test and post-test sessions, participants in both the training and control groups completed assessments regarding LD knowledge, teaching efficacy and attitudes towards students with LDs. Each participant required approximately 40 min to complete all the tests in each session. The pre-test and post-test sessions were conducted within one week prior to and following the training period, respectively. During the intervention period, participants in the training group received LD knowledge training in addition to their regular academic activities, whereas the control group continued with their usual instructional activities without any additional intervention. Consistent with the procedure in relevant literature (Alabd et al. 2018; Latouche and Gascoigne 2019), the LD knowledge training spanned 4 weeks and comprised 4 training sessions (one session per week), with each session lasting approximately 30 min. The whole training program was administered by an experimenter who had received professional training. Written informed consent was obtained from the participants before the assessment.

2.3. Assessments in the Pre- and Post-Tests

Knowledge of learning disabilities. The LD knowledge test was developed based on the content of Chapter 1 “Special Pedagogy” and Chapter 13 “Education of children with disabilities” in the book of Special Education (Fang 2005). In order to minimize practice effects, two parallel versions of the test (i.e., Version A and Version B) were created. Students who majored in psychology were invited to conduct a subjective assessment of the homogeneity between the two sets of items. Results showed no significant difference in difficulty or content structure between the two versions (ps > 0.1). The test covered the definition, characteristics, and relevant teaching strategies regarding LDs. It consisted of 29 judgment questions, with each correct answer awarded 1 point, resulting in a maximum score of 29. Higher total scores indicated a more comprehensive grasp of LD concepts among pre-service teachers.

Teaching efficacy. The teaching efficacy scale developed by Tschannen-Moran and Hoy (2001) was used here. This scale consists of three dimensions: student engagement, instructional strategies, and classroom management. There are a total of twelve items rated on a 5-point Likert scale (i.e., one point represents the teacher’s feeling of being completely incompetent, five points represent that the teacher feels completely competent). The scores for all items were summed to generate a final score, where higher values corresponded to stronger teaching efficacy. The scale demonstrates robust psychometric properties. The reliability of this scale is 0.9, with internal consistency coefficients for the three dimensions ranging from 0.81 to 0.86, reflecting satisfactory validity and reliability (Tschannen-Moran and Hoy 2001).

Teachers’ attitudes. The teacher attribution questionnaires were adapted from the study of Clark (1997). The original instrument comprised eight vignettes that examined how American primary school teachers perceive the academic achievements of students with and without specific learning disabilities. For the present study, we selected the four vignettes specifically depicting hypothetical male students with LDs who had failed a classroom test. Male protagonists were retained to control for the potential confounding variable of biological sex, and the textual content was refined to align with the cultural context of mainland China. Following each vignette, two 7-point Likert scale items measured the teacher’s anger towards the student’s test failure (e.g., “Would Jim make you feel angry?” where 1 indicates “not at all angry” and 7 indicates “extremely angry”) and their cognitive expectation of future failure (e.g., “How likely is it that Jim will fail the next exam?” where 1 indicates “very unlikely” and 7 indicates “very likely”). To ensure methodological rigor, the psychometric properties of the adapted instrument were re-evaluated. Reliability analyses indicated acceptable internal consistency for both the emotional dimension (Cronbach’s Alpha = 0.64) and the cognitive dimension (Cronbach’s Alpha = 0.60). Construct validity was examined by analyzing the correlation between the two dimensions. Teachers’ anger was significantly negatively correlated with their expectations of future failure (r = −0.45, p < 0.001), indicating that reduced anger corresponded to a higher perceived likelihood of future failure for students with LDs. This suggests that the two dimensions converge well in measuring the underlying construct of teacher attribution. Furthermore, content validity is supported by the instrument’s theoretical basis in Clark’s (1997) research. The mean score across the four vignettes was computed to derive composite scores for the emotional and cognitive dimensions of teacher attributions, while the order of presentation was randomized across participants to mitigate sequence effects.

2.4. Training of Learning Disability Knowledge

The training curriculum encompassed four key modules: (1) the definition and classification of LDs; (2) the possible causes, assessment, and identification of LDs; (3) the psychological characteristics of children with LDs; (4) educational strategies and supports for students with LDs. The training materials were developed based on pivotal chapters from authoritative textbooks such as ‘Special Education’ (Fang 2005). The detailed training schedule is outlined in

Table 1. Detailed information about the LD knowledge training program.

Theme	Content
Session 1: Definition and classification of LD	General introduction of the concept, classification, principles and process of special education What do you think about learning disabilities (Discussion + video materials) Definition of LD (Interaction) Classification of LD
Session 2: Possible causes, assessment, and identification of LD	Causes of learning disabilities (Discussion + video materials) Scenario Reenactment—Arbitrary Labeling of Children with Learning Disabilities by Teachers and Parents Assessment and identification of LD
Session 3: Psychological characteristics of children with LD	Cognitive processing skills of children with LD Mental health of children with LD Impact factors of psychological development of children with LD Demonstration case studies in teaching children with LD (Discussion + video materials)
Session 4: Educational strategies and supports for students with LD.	Introduction of cases of teacher training program Direct Instruction Model Interactive and Heuristic Teaching Model Approaching children with learning disabilities from a developmental perspective (Discussion + video materials)

Note. LD: learning disability.

.

The training was delivered online to enhance implementation feasibility. To ensure effective learning, each course session incorporated interactive elements such as discussions and reflective activities. To further support optimal learning outcomes, pre-service teachers were provided with comprehensive learning materials (including handouts, presentations, and supplementary resources) one day prior to each session to facilitate preview. Following this, the course videos were released, requiring participants to complete their online learning within a designated week and check-in to confirm progress. After logging their completion, participants were issued after-class assignments, which they were required to submit promptly before being granted access to the next session. Participants in both the training and control groups were informed at the beginning of the training that there would be a test on LD knowledge upon completion of the program. They were advised that only those achieving a passing score would receive incentives, thereby motivating all participants especially those in the training group to engage deeply with the online course.

2.5. Plans of Data Analyses

Firstly, to examine the direct and transfer effects of the training intervention, a series of 2 (time point: pre- vs. post-tests) × 2 (group: training vs. control group) repeated measure analysis of variances were separately conducted on LD knowledge, teaching efficacy, pre-teachers’ attitudes towards academic failure of students with LD. Then, to further investigate the relationship among potential training effects, change scores were computed by subtracting pre-test scores from post-test scores for each variable. Pearson correlation analyses were then performed to explore the association between changes in LD knowledge, teaching efficacy, and teacher attitudes. Subsequently, mediation analyses were conducted using the PROCESS macro for SPSS (Hayes 2013). The analyses utilized a resampling strategy of 1000 bootstraps to generate bias-corrected 95% confidence intervals (CIs). An indirect effect was considered statistically significant (p < 0.05) if the confidence interval did not include zero. All analyses were performed using SPSS version 27.0.

3. Results

3.1. Training Effects on LD Knowledge, Teaching Efficacy, and Teachers’ Attitudes

3.1.1. The Pre-Post Comparison of LD Knowledge

Results of the repeated measures ANOVA indicated that the main effect of time point was significant [F(1,49) = 181.82, p < 0.001, η² = 0.79], that is, the pre-service teachers’ LD knowledge significantly increased from the pre-test to the post-test. The main effect of group was not significant [F(1,49) = 2.72, p = 0.11, η² = 0.05]. However, the interaction between time point and group was significant [F(1,49) = 42.80, p < 0.001, η² = 0.47]. Results of simple effect analyses (

Table 2. Comparisons of pre- and post-test scores in training and control groups.

Items	Intervention Group (N = 26)		Control Group (N = 25)
	Pretest	Post-Test	Pretest	Post-Test
LD knowledge	18.27 ± 2.03	24.96 ± 2.46	19.68 ± 0.90	22.00 ± 2.43
Teaching efficacy	3.52 ± 0.34	3.90 ± 0.56	3.74 ± 0.23	3.82 ± 0.24
Teachers’ attitudes to academic failure of LD cases
Dimension of emotional arousal	3.31 ± 0.84	2.51 ± 0.96	3.46 ± 0.87	3.54 ± 0.90
Dimension of cognitive evaluation	4.16 ± 0.74	4.16 ± 0.70	4.16 ± 0.83	4.32 ± 0.79

Note. LD: learning disability.

) showed that the pre-service teachers in both the training [F(1,49) = 204.53, p < 0.001, η² = 0.81] and control [F(1,49) = 23.63, p < 0.001, η² = 0.33] groups had significantly higher LD knowledge levels at the post-test than those at the pre-test. In the pre-test, the LD knowledge level was significantly lower in the training group than in the control group [F(1,49) = 10.14, p = 0.003, η² = 0.17]; while in the post-test, the training group demonstrated significantly higher LD knowledge level than the control group [F(1,49) = 18.70, p < 0.001, η² = 0.276]. Furthermore, change scores in LD knowledge between groups were submitted to the independent samples t test. The result indicated that the improvement in the training group was significantly greater than that in the control group (t = 6.54, p < 0.001).

3.1.2. The Pre-Post Comparison of Teaching Efficacy

Results of the repeated-measures ANOVA similarly as above showed that the main effect of time point was significant [F(1,49) = 9.07, p = 0.004, η² = 0.16], with higher teaching efficacy of pre-service teachers at the post-test stage as compared to that at the pre-test stage. The main effect of group was not significant [F(1,49) = 0.94, p = 0.34, η² = 0.02], yet the interaction between time point and group was marginally significant [F(1,49) = 3.81, p = 0.06, η² = 0.08]. Further results of the simple effect analysis (Table 2) revealed that teaching efficacy in the training group improved significantly from pre-test to post-test [F(1,49) = 12.56, p < 0.001, η² = 0.20]; whereas the teaching efficacy of the control group did not differ significantly between the two time points [F(1,49) = 0.55, p = 0.46, η² = 0.01]. At the pre-test stage, teaching efficacy in the training group was significantly lower than that of the control group [F(1,49) = 7.02, p = 0.011, η² = 0.13], with no significant difference between the training and the control groups at the post-test stage [F(1,49) = 0.44, p = 0.51, η² = 0.01].

3.1.3. The Pre-Post Comparison of Teachers’ Attitudes

Dimension of emotional arousal. The results of the repeated-measures ANOVA on pre-service teachers’ anger arousal in response to the academic failure of students with LD showed a significant main effect of time point [F(1,49) = 4.85, p = 0.03, η² = 0.09]. This indicates that pre-service teachers experienced less anger regarding the academic failure of students with LD at the post-test stage compared to the pre-test stage. The main effect of group was also significant [F(1,49) = 9.61, p = 0.003, η² = 0.16], with the control group reporting higher anger level than the training group. The time point × group interaction was significant [F(1,49) = 7.26, p < 0.01, η² = 0.13]. Results of the simple effect analysis (Table 2) indicated that anger evoked by the academic failure of students with LD decreased significantly in the training group from pre-test to post-test [F(1,49) = 12.231, p < 0.01, η² = 0.20]. In contrast, the pre-service teachers in the control group did not change significantly in their anger over academic failure of the students with LD between the pre- and post- tests [F(1,49) = 0.12, p = 0.73, η² = 0.002]. While the group difference was not significant at the pre-test [F(1,49) = 0.40, p = 0.53, η² = 0.01], the training group reported significantly lower anger compared to the control group at the post-test [F(1,49) = 15.55, p < 0.001, η² = 0.24].

Dimension of cognitive evaluation. Pre-service teachers’ cognitive evaluation of future academic failure for the LD cases was taken as the dependent variable, which was submitted to the same ANOVA as above. Results showed no significant main effects or interaction (ps > 0.1).

3.2. Relations of Training Changes Across LD Knowledge, Teaching Efficacy, and Teachers’ Attitudes

The subtractions of scores between post- and pre- tests were separately computed as the training changes in LD knowledge levels, teaching efficacy, and teacher’s attitudes (including anger arousal to the academic failure of the LD cases and cognitive evaluation of future academic failure for the LD cases) in the training group. Correlation analyses (

Table 3. Correlations of the training changes across LD knowledge, teaching efficacy, and teachers’ attitudes.

Post-Pre Comparisons	Mean	Standard Deviation	1	2	3
1 Increased LD knowledge	6.69	2.59	-
2 Enhanced teaching efficacy	0.38	0.68	0.03	-
3 Decreased anger arousal to students’ academic failure	−0.80	1.41	−0.48 *	−0.45 *	-
4 Changes in cognitive evaluation of students’ future academic failure	<0.001	0.73	0.18	−0.50 *	0.35

Note. LD: learning disability. Training changes are the subtractions of scores between post- and pre- tests. *, p < 0.05.

) were conducted to examine the relationships among those training changes. Results showed that the training change in anger arousal was negatively correlated with changes in both LD knowledge (r = −0.48, p = 0.013) and teaching efficacy (r = −0.45, p = 0.02). This indicates that a greater decrease in pre-service teachers’ anger arousal regarding the academic failure of students with LD corresponded with greater gains in their LD knowledge and teaching efficacy. Moreover, the training change in pre-service teahcers’ cognitive evaluation regarding the future failure of LD cases was negatively correlated with the improvement in their teaching efficacy (r = −0.50, p < 0.01). This indicates that a lower perceived likelihood of future failure of LD cases corresponded with greater improvement in teaching efficacy. There were no other significant correlations (ps > 0.1).

Considering that the training change in pre-service teachers’ anger arousals to the academic failure of students with LD were associated with both gains in LD knowledge and teaching efficacy, mediation analyses were conducted. These analyses examined whether improvements in LD knowledge directly contributed to training benefit in teaching efficacy, or if this relationship was mediated by changes in teacher’s attitude (i.e., anger arousal to the academic failure of students with LD and cognitive evaluation of future academic failure).

As shown in

Table 4. Mediation analyses on training changes in LD knowledge and teachers’ attitudes, and teaching efficacy.

Factors	F Values	∆R2	SE (Boot)	95% CI
Increased LD knowledge→Decreased anger arousal→Enhanced teaching efficacy
Total effect	0.02	0.1%	0.05	[−0.1029, 0.1193]
Direct effect	3.78	24.76%	0.05	[−0.1759, 0.0491]
Indirect effect	-	-	0.04	[0.0013, 0.1367]
LD→AA	7.18 *	23.02%	0.10	[−0.4611, −0.0598]
AA→TE	6.12 *	20.3%	0.10	[−0.4821, −0.0678]
Increased LD knowledge→Improved cognitive evaluation→Enhanced teaching efficacy
Total effect	0.02	0.1%	0.05	[−0.1029, 0.1193]
Direct effect	3.78	24.76%	0.05	[−0.0657, 0.1323]
Indirect effect	-	-	0.03	[−0.0897, 0.0275]
LD→CEFAF	0.82	3.29%	0.53	[−0.0652, 0.1667]
CEFAF→TE	4.21 *	26.78%	0.37	[−0.8486, −0.1412]

Note. LD: learning disability knowledge; AA: anger arousal; TE: teaching efficacy; CEFAF: cognitive evaluation of future academic failure. SE: standard error. CI: confidence interval. *, p < 0.05.

and Figure 2, the results exhibited that the total effect of improved LD knowledge on the training changes in teaching efficacy was not significant (βc = 0.0082, 95% CI: [−0.1029, 0.1193]). While increased LD knowledge showed no direct effect on enhancing teaching efficacy (βc’ = −0.063, 95% CI: [−0.1759, 0.0491]), a significant indirect effect was observed (β = 0.0716, 95% CI: [0.0013, 0.1367]), mediated by training-induced changes in anger arousal evoked by the academic failure of LD cases.

In addition, when using changes in cognitive evaluation of academic failure as the mediator, increased LD knowledge showed neither a direct effect (βc’ = 0.0333, 95% CI: [−0.0657, 0.1323]) nor an indirect effect (β = −0.0251, 95% CI: [−0.0897, 0.0275]) on teaching efficacy.

4. Discussion

The present study designed and evaluated a training program aimed at enhancing pre-service teachers’ knowledge of LDs. The results indicated that the training significantly improved the LD knowledge of the pre-service teachers. Specifically, the mean accuracy score on the LD knowledge test escalated from 63% (i.e., 18 correct answers out of 29) at the pre-test phase to 86% (i.e., 25 correct answers out of 29) post-training. This baseline LD knowledge level observed in our sample aligns with prior research (Yin et al. 2020), which reported a mean accuracy of 58% on an LD knowledge test among primary school educators in economically advanced regions of mainland China. Similarly, the pre-service teachers involved in this study, primarily aspiring to teach in K-12 settings upon graduation, hailed from developed regions within mainland China. These findings underscore that, despite originating from more affluent regions, the initial LD knowledge levels of Chinese pre-service teachers remain comparatively lower than those of their counterparts in other countries, where reported accuracies stood at 77.9% and 77.43%, respectively (Aguiar et al. 2014; Rajesh and Joseph 2023). Consequently, the present study implemented a targeted training intervention, achieving a substantial improvement of 36.5% in LD knowledge from pre-test to post-test, thereby bringing the knowledge levels closer to those observed in developed nations, where such improvements typically range around 19% (Aguiar et al. 2014). This marked enhancement underscores the efficacy of the implemented training program.

Furthermore, the current training regarding LD knowledge exerted significant transfer effects, by enhancing pre-service teachers’ teaching efficacy and attenuating their anger arousal in response to academic failures experienced by students with LD. Crucially, mediation analysis revealed that enhanced LD knowledge among pre-service teachers predicted improved teaching efficacy, and this relationship was mediated by a reduction in anger arousal. This relationship pattern across LD knowledge, teachers’ attitudes, and teaching efficacy from the dynamic aspect was consistent with the prediction proposed on the basis of the multidimensional model of professional teaching competencies (Baumert and Kunter 2013). Analysis of the training content and the LD knowledge test items indicated that the gains in LD knowledge were predominantly in the areas of definition, identification, and diagnosis of LD. We infer that this acquisition of LD knowledge facilitates a process of cognitive reinterpretation regarding students’ academic failure. When pre-service teachers can attribute the academic shortcomings of students with LD to neurodevelopmental factors rather than motivational deficits, they are less likely to perceive failure as a personal affront or a sign of student defiance. Consequently, this objective reframing mitigates the elicitation of negative emotions such as anger (Gross 1999, 2002). Furthermore, in accordance with the premise that affective states serve as informational cues for self-efficacy judgments (Tschannen-Moran et al. 1998), negative emotions elicited by students’ failure experiences can undermine teachers’ teaching efficacy (Burić et al. 2020). Conversely, the attenuation of negative emotion such as anger functions as a protective mechanism regarding teaching efficacy. By reducing the emotional toll associated with teaching challenging cases, the improvement of LD knowledge can offset potential declines in self-perceived competence. The above findings delineate a plausible pathway linking teachers’ cognitive characteristic (i.e., LD knowledge) with affective/motivational characteristic (i.e., teachers’ attitudes). Specifically, enhanced LD knowledge reduces the anger arousal stemming from misinterpreting LD-related failures, thereby preserving and enhancing teaching efficacy. Sustaining high teaching efficacy is imperative, as it directly influences the accuracy of LD identification and the willingness to initiate supportive interventions (Kirby et al. 2005; Sciutto et al. 2000).

It is noteworthy that the LD knowledge training selectively influenced the emotional dimension of pre-service teachers’ attitudes toward the academic failure of students with LD while leaving the cognitive dimension unchanged. Despite the training, pre-service teachers in the current study continued to hold the belief that students with LD were more likely to experience future learning difficulties. Consequently, this static cognitive dimension of attitude failed to mediate the relationship between increased LD knowledge and the improvement in teaching efficacy. Furthermore, the direct effect of the acquisition of LD knowledge on the enhancement of teaching efficacy was not observed in the current study, which was inconsistent with previous research reporting that increased LD knowledge boosted teaching efficacy (Alabd et al. 2018; Jarque Fernández et al. 2021; Latouche and Gascoigne 2019). These null findings were potentially due to a misalignment between the training’s theoretical focus and the participants’ lack of professional context. The current training focused on the definition, identification, and diagnosis of LD. Although educational strategies for children with learning disabilities were included, the instruction remained heavily theoretical, lacking components that facilitate the application of theory to practice. Meanwhile, it is important to note that the pre-service teachers possessed limited teaching experience compared to the in-service teachers examined in previous research (Ismailos et al. 2019), making it difficult for them to integrate abstract theoretical knowledge without a practical framework (Ohst et al. 2015). This lack of a practical context implies that the training may have failed to provide sufficient enactive mastery experiences regarding LD instruction. While the training may have fostered vicarious experiences and influenced physiological/affective states regarding LD cases, mastery experience remains the primary information source of self-efficacy (Lippke 2017). Given its pivotal role, the absence of such enactive mastery experiences meant that the newly acquired LD knowledge could not be effectively translated into robust teaching efficacy, thereby failing to establish this specific pathway from the increased LD knowledge to improved teaching efficacy. Adequate teaching experience is essential for the deeper comprehension and internalization of training content, facilitating the translation of pedagogical knowledge into effective classroom practice. Applying LD knowledge to instruction is vital for enhancing the academic performance of students with LD and reducing the probability of future academic failure. Therefore, future iterations of the LD knowledge training should move beyond theoretical definitions to incorporate specific instructional strategies and practical sessions for pre-service teachers. This approach would help increase the enactive mastery experience necessary to effectively boost teaching efficacy. By emphasizing the application of LD knowledge to daily instructional tasks, such training can better support the genuine improvement of the academic outcomes of students with LD.

There are several limitations to the present study that warrant acknowledgment. Firstly, the study measured general teaching efficacy rather than efficacy specifically targeting the instruction of students with learning disabilities. This discrepancy may explain the lack of a direct link between LD knowledge and teaching efficacy. Future research should investigate domain-specific teaching efficacy to better understand the contributions of LD knowledge training. Secondly, the incentive structure of the experiment, where all participants were informed that only those who passed the post-test would receive incentives, may have introduced a confounding variable of extrinsic motivation. This condition might have encouraged autonomous learning or strategic guessing among participants in both the training and control groups. While the equal application of this incentive between groups and our focus on the interaction effect in the statistical analysis help to control for this bias, it remains a potential influence on the observed outcomes. Thirdly, although this study randomly assigned pre-service teachers with low LD knowledge and low teaching efficacy into training and control groups, a significant baseline difference in these measures was observed between the two groups. While the use of 2 (Group: training vs. control groups) × 2 (Time: pre- vs. post-tests) repeated-measures ANOVAs statistically controlled for the influence of these pre-test differences on the analysis of training effects, future research should further confirm the specific impact of the training after ensuring balanced baseline levels. Fourthly, the LD knowledge tests (Version A and B) used in this study were self-developed and based on existing special education tests. However, no large-scale predictive validity testing was conducted, leaving the reliability and validity of these instruments unclear. Fifthly, the duration of the training program was relatively brief, consisting of only four sessions totaling approximately two hours of direct instruction. While this design was informed by previous research (e.g., Alabd et al. 2018; Latouche and Gascoigne 2019) indicating the efficacy of short-term training for improving specific knowledge domains, the brevity of the program may limit its capacity to induce profound or long-lasting changes in complex attitudinal constructs. Consequently, the generalizability of the findings to longer-term training programs remains to be established. Future research should therefore extend training periods to solidify these changes. Furthermore, learning disabilities constitute a heterogeneous category, and the current training did not differentiate instructional strategies based on specific subtypes (e.g., dyslexia, dyscalculia, or ADHD). While the session included examples illustrating the distinct profiles of these subtypes, the training primarily focused on general LD knowledge rather than tailored intervention techniques. Given the varying behavioral manifestations and classroom management needs of students with different LD subtypes, future research should develop and evaluate LD-specific training modules. Extending the training duration would allow educators to acquire targeted practical skills for addressing the unique challenges posed by specific learning difficulties, thereby fostering more nuanced and effective support. Sixth, it is important to recognize that attitudes are complex, multi-dimensional, and relatively stable constructs. Previous literature suggests that even when declarative attitudes appear to improve following an intervention, behaviors may persist due to the gap between cognition and action (e.g., Conner and Norman 2026; Costa 2024). In the present study, attitudes were measured exclusively through responses to hypothetical vignettes, which may not fully capture the nuances of real-world classroom interactions. Future research should therefore incorporate behavioral indicators such as classroom observations or teacher-student interaction analyses, so as to more comprehensively examine the potential impact of LD knowledge enhancement on actual teaching practices and behavioral responses toward students with LD. Lastly, the current study recruited pre-service teachers who lack deeply entrenched pedagogical beliefs and are likely in a critical phase of professional socialization. This contrasts sharply with in-service teachers, who often require the deconstruction of established beliefs. Consequently, the observed improvements in attitudes here may be more accurately conceptualized as attitude formation rather than attitude change. Future research should therefore investigate whether the mechanisms driving attitude formation in novices differ from those required for attitude change in experienced educators. Addressing these methodological limitations will be crucial for strengthening the robustness of future research findings.

5. Conclusions

In summary, the present study designed and implemented a training program to enhance LD knowledge among pre-service teachers. The findings indicate that the training significantly improved participants’ LD knowledge, domain-general teaching efficacy, and reduced anger arousal in response to the academic failures of students with LD. Furthermore, the results revealed a complex, triangular relationship among these factors. Specifically, the enhancement of pre-service teachers’ LD knowledge predicted an increase in their general teaching efficacy, mediated by changes in the emotional dimension of their attitudes, namely, a decrease in their anger arousal towards the academic failures of students with LD. These findings suggest that equipping pre-service teachers with specialized LD knowledge positively impacts their overall teaching efficacy. This relationship appears to be partially mediated by a reduction in pre-service teachers’ negative emotional responses when confronted with the academic challenges faced by students with learning needs. Consequently, these findings underscore the importance of integrating targeted LD training into pre-service education programs to foster comprehensive teaching competencies among future educators and promote a more inclusive educational environment for students with learning disabilities.